1 |
On Wed, Nov 04, 2009 at 11:12:05PM +0100, Peter Hjalmarsson wrote: |
2 |
> tis 2009-11-03 klockan 16:48 +0100 skrev Patrick Lauer: |
3 |
> > Hi there, |
4 |
> > |
5 |
> > All of these bugs were for the use of the FEATURES variable in ebuilds, which |
6 |
> > is a very convenient thing to work around issues. |
7 |
> > For example known failures with FEATURE="distcc" or funky things like test |
8 |
> > failures with FEATURES="userpriv" and so on. All other methods of expressing |
9 |
> > that are much more verbose and inherently sucky. |
10 |
> |
11 |
> I ask myself if what we really want is many different and strange |
12 |
> approaches to handle FEATURES? |
13 |
> Would it not be better to actually expand some eclasses to be able to |
14 |
> say something about your build environment? |
15 |
|
16 |
This is a *really* bad approach- pkgcore/paludis have supported |
17 |
standalone repositories basically from their inception, portage (via |
18 |
repos.conf) has developed equivalent support, and from the looks of it |
19 |
is moving towards such capabilities. |
20 |
|
21 |
What this means is that eclasses aren't automatically mashed together |
22 |
across all trees and shared. Meaning each repository would have to |
23 |
carry those eclasses, or require that they always be slaved against a |
24 |
specific repository carrying said eclasses (again, bad mojo). |
25 |
|
26 |
The code duplication there is annoying, but the potential range of |
27 |
screwups this can lead to is even worse. Further, via proper |
28 |
environment saving/restoration for installed pkgs/binpkgs, this means |
29 |
that one screwed up FEATURES detection (or that things have changed |
30 |
since then and a new check is required) lives on forever, no |
31 |
possibility of being updated/fixed in the env dump. |
32 |
|
33 |
Shorter version, eclasses are a fun idea at first until you dig in and |
34 |
realize they'll blow your leg off for things like this. |
35 |
|
36 |
If it's any consolation, I proposed moving large chunks of eapi0 |
37 |
(prior to eapi existing) into the tree- the idea died off for the same |
38 |
reasons your "FEATURE awareness in eclasses" must die off. |
39 |
|
40 |
~harring |